** readMe **

This folder contains 10 imputed data files, each of which must be analyzed (and the results aggregated) in order to replicate the results in main text of the article. 

Replication should be carried out in the following order:

	1. Estimate the mixed logit models in the Stata do file called "estimation.do" 
		1a. Feel free to reduce the number of posterior to draws to as low as 100 if you wish
		to save time simulating the electoral losses, but note that your results will be much 
		less likely to match those in the article
	2. Run through the R script called "replication.r" to generate the main table and figures
	3. Execute each of the 10 "sim.r" scripts to simulate the electoral losses resulting from a change in perceived compromise.
		3a. These can (and should) be batched in parallel.
	4. Run through the R script called "secondAnalysis.r" to generate the results in Table 3

**##!!! DO NOT FORGET TO PROPERLY SET ALL DIRECTORIES IN ALL SCRIPTS !!!##**